Full Event Particle-Level Unfolding with Variable-Length Latent Variational Diffusion
This addresses a bottleneck in particle physics for researchers needing accurate detector effect corrections in variable-dimensional environments, representing a novel method rather than an incremental improvement.
The paper tackles the problem of full-event particle-level unfolding in variable-dimensional collider data by introducing a novel modification to variational latent diffusion models, enabling unfolding of high- and variable-dimensional feature spaces, with performance evaluated in semi-leptonic top quark pair production at the Large Hadron Collider.
The measurements performed by particle physics experiments must account for the imperfect response of the detectors used to observe the interactions. One approach, unfolding, statistically adjusts the experimental data for detector effects. Recently, generative machine learning models have shown promise for performing unbinned unfolding in a high number of dimensions. However, all current generative approaches are limited to unfolding a fixed set of observables, making them unable to perform full-event unfolding in the variable dimensional environment of collider data. A novel modification to the variational latent diffusion model (VLD) approach to generative unfolding is presented, which allows for unfolding of high- and variable-dimensional feature spaces. The performance of this method is evaluated in the context of semi-leptonic top quark pair production at the Large Hadron Collider.